1. Family planning programs are so new that the data available barely cover a sufficient length of time to support any kind of rate-of-return analysis. Rates of return based on the early years of a program may bear little relationship to subsequent rates of return. In the early years economies of scale and also of experience have not been fully reaped; and on the other hand, it may have been possible to "skim the crearn" by making acceptors out of those women who are strongly motivated to practice family planning.
2. The major "output" from which rates of return can be computed is the number of births averted by the family planning program, but this will always be difficult to measure. There is enough well-founded evidence to be confident that family planning programs have had, in some cases at least, an independent effect on the birth rate.* But measurement of this effect is normally subject to a wide range of uncertainty.!
3. A further problem, related to both the foregoing points, is that the potential, as well as actual, impact of a family planning program on the birth rate may change continuously over time. Change in the socioeconomic matrix, in which the family planning program is one element, may after a time render the program obsolete as a method of bringing down the birth rate.
4. There is continuing controversy over the correct rate at which the future consumption and production streams generated by an as yet unborn individual are to be discounted (76). A high rate of discount will "wash out" the production effect almost completely since this does not begin for at least a decade. Other crude aspects of the usual "cost-benefit" analysis of birth prevention, such as the assumption that additional newborn infants will consume at the same rate as the average member of the population they enter, and the tendency to reckon private costs to the public account, have also been questioned (59, 77, 78, 79).
*This evidence is of a variety of kinds. For example, in Taiwan, the rate of fertility decline has accelerated markedly following the introduction of the family planning program in 1963, and both the percent of wives 20 to 44 who had ever practiced contraception and of those who were currently practicing increased by 50 percent in 2 years-1965-67 (60, No. 41, Fig. 7; 72).
A "matching study" which compared the fertility of women accepting intra-uterine devices (lUD's) in the Taiwan program with that of nonacccptors with similar characteristics indicates that the fertility of acceptors fell by 80 percent, whereas that of nonacceptors fell by 48 percent (73).
In Korea the strongest evidence of program impact is the remarkable pace of change in rural areas. The highest acceptance rate for contraception has been among rural women with little or no education, persons who would be the least likely to practice contraception in the absence of the program. The percentage of married women through age 49 currently using contraception in rural areas rose from 6 percent in 1964 (the year the family planning program was initiated) to 18 percent in 1966 (74).
TFor a concise survey of methods for estimating births averted by family planning programs, see (75).